Simulated wheat growth affected by rising temperature, increased water deficit and elevated atmospheric CO2

The cropping systems simulation model APSIM-Nwheat was tested against detailed field measurements representing possible growing conditions under future climate change scenarios. Increasing average temperatures by 1.7 degreesC observed over several seasons at Obregon, Mexico reduced the time to flowering by 11 days and resulted in a decline of total biomass and grain yield. These effects were reproduced by the model, except when the observed total biomass inexplicably rose again in the fourth and fifth year, despite higher temperature and a much shorter growing time. In a water stress experiment, the effects of different timing and duration of water deficit on crop growth and yield were reproduced with the model for a rain-shelter experiment at Lincoln, New Zealand where observed grain yields were reduced from 10 to 4 t ha(-1) due to increased water deficit. In experiments from Western Australia, reduced growth and yields due to extreme terminal water deficit were also reproduced with the model where measured yields fall below 0.5 t ha(-1). In the Maricopa Free Air Carbon-Dioxide Enrichment (FACE) experiment in Arizona, USA, the largest yield increase occurred with elevated CO2 in the dry and high N treatments, whereas little or no response was observed in the wet and low N supply treatments, as simulated with the model. Combining elevated CO2 with increased temperature in a sensitivity analysis, two levels of water supply and a range of N applications indicated a positive effect of elevated CO2 on yield as long as N was not limiting growth. Increased temperature and reduced water supply reduced yields and the yield response to N supply under ambient and elevated CO2. Grain protein concentrations were reduced under elevated CO2, but the difference was minor with ample N fertiliser. Evapotranspiration was reduced under elevated CO2. Higher temperatures increased evapotranspiration with low N input, but reduced it with ample N fertiliser, resulting in a reduction and an increase, respectively, in drainage below the root zone. In the Mediterranean environment of Western Australia the impact of elevated CO2 and increased temperature on grain yield was in average positive, but varied with seasonal rainfall distribution. Based on the range of model testing experiments and the sensitivity analysis, APSIM-Nwheat was found suitable for studies on directional impacts of future climate change on wheat production. Due to some large discrepancies between simulated and observed data, field experiments representing only a limited range of possible climate change scenarios and the large possible range of factorial interactions not tested, simulated quantitative effects with the model should be interpreted cautiously. (C) 2003 Elsevier B.V. All rights reserved.